Dermal absorption of cyclic and linear siloxanes: a review

ABSTRACT

Cyclic and linear siloxanes are compounds synthesized from silicon consisting of alternating atoms of silicone and oxygen [Si-O] units with organic side chains. The most common cyclic siloxanes are octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6), while the most common linear siloxanes are high molecular weight polydimethylsiloxanes (PDMS) and low molecular weight volatile linear siloxanes known as hexamethyldisiloxane (L2), octamethyltrisiloxane (L3), decamethyltetrasiloxane (L4), dodecamethylpentasiloxane (L5). These compounds (1) exhibit low dermal toxicity, (2) are generally inert and non-reactive, and (3) are compatible with a wide range of chemicals offering beneficial chemical properties which include the following: wash-off or transfer resistance from the skin, sun protection factor (SPF) enhancement, emolliency in cleaning products). Because of these properties, these compounds are incorporated into multiple consumer products for use on the skin, such as cosmetics and health-care products, with over 300,000 tons annually sold into the personal care and consumer products sector. Because of their widespread use in consumer products and potential for human dermal exposure, a comprehensive understanding of the dermal absorption and overall fate of siloxanes following dermal exposure is important. This review summarizes available data associated with the dermal absorption/penetration as well as fate of the most commonly used siloxane substances.

KEYWORDS:

• Siloxanes
• dermal
• absorption
• linear
• cyclic
• cosmetics
• sunscreen products

Introduction

The structure of siloxanes is represented by repeating groups of –[R₂Si-O]- with organic substitutions at the silicon atom (Huber and Kaiser 1986). The substituent R is usually represented as methyl, ethyl, propyl, phenyl, fluoroalkyl, aminoalkyl, hydroxy, mercaptoalkyl, hydrogen or vinyl groups. Based upon chemical structure, siloxanes may be distinguished as linear, cyclic, branched or crosslinked. Siloxanes, both cyclic and linear (low molecular weight) and linear (high molecular weight) compounds are synthesized from silicon (Si) and consist of alternating atoms of silicone and oxygen [Si-O] units with organic side chains (Horii and Kannan 2008; Mojsiewicz-Pieńkowska et al. 2016).

Due to the overall low dermal toxicity of siloxanes, these chemicals are generally inert, non-reactive and compatible with a wide range of chemicals. These compounds provide numerous benefits including skin sensory and texture enhancement, emolliency and spreadability, transient to long-lasting effects, wash-off or transfer resistance from the skin, nonocclusive to semipermeable, sun protection factor (SPF) enhancement, protection and cleaning of the skin and thus employed in consumer products for use on the skin in cosmetics and health-care products (Horii and Kannan 2008; Mojsiewicz-Pieńkowska and Krenczkowska 2018; Mojsiewicz-Pieńkowska et al. 2016).

Cyclic siloxanes are individual-chain length cyclic dimethyl polysiloxane compounds (Johnson et al. 2011; Krenczkowska et al. 2020; Wang et al. 2009). The most common cyclic siloxanes are octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6) (Johnson et al. 2011). D4, D5, and D6 consist of 4, 5, or 6 repeating units of [(CH₃)₂SiO] connected in a ring attachment (Jovanovic et al. 2008). Linear siloxanes, the most common of which are polydimethylsiloxanes (PDMS), are viscous polymeric silicone compounds composed of repeating units of [(CH₃)₂SiO] with viscosities ranging from 10 to > 100,000 centistokes (cSt) (ECETOC 2011). In addition to high molecular weight PDMS, low molecular weight volatile linear siloxanes known as hexamethyldisiloxane (L2), octamethyltrisiloxane (L3), decamethyltetrasiloxane (L4), dodecamethylpentasiloxane (L5) might also be present in cosmetics, and medical devices (GSC 2023).

This review specifically focused on the dermal absorption potential of cyclic and linear (low and high molecular weight) siloxanes. The molecular weight and chemical structure of siloxanes determines the physiochemical properties such as solubility, lipophilicity, and volatility, ability to permeate through skin, capacity to overcome cellular barriers, and toxicity. Both linear and cyclic siloxanes display distinct chemical and physical properties including inertness, reliable performance at high and low temperatures, water repellency and low surface tension, making them attractive for use in a wide range of industrial, pharmaceutical and medicinal applications, as well as dermatological, cosmetic, personal care, and household items (ECETOC 2011; Johnson et al. 2011; Jovanovic et al. 2008).

According to the Global Silicones Council (GSC 2023) 390,000 tons of silicone products are sold into the personal care and consumer products sector each year and are used in a wide range of products. Horii and Kannan (2008) determined the concentrations of cyclic and linear siloxanes in a variety of consumer products and found that concentrations of cyclic siloxanes in consumer products ranged from < 0.35 to 9380 µg/g, <0.39 to 81,800 µg/g, and < 0.33 to 43,100 µg/g for D4, D5, and D6, respectively, while concentrations of linear siloxanes (L4 to L14) varied from < 0.059 to 73,000 µg/g. More than 50% of the samples analyzed by Horii and Kannan (2008) contained D4, D5, or D6. In their survey and risk assessment of siloxanes in cosmetic products, the Danish Environmental Protection Agency (2021) selected 6 siloxanes including D4, D5 and D6, for quantitative analysis in cosmetic products. Twenty-five of the 40 cosmetic products analyzed contained either a single component of D4, D5, or D5 or a mixture of the three cyclic siloxanes. The products containing cyclic siloxanes included facial and body creams, deodorants, foundation hair balm, hair sprays, hair oils, hair shampoos, sun creams and sun sprays. Results of the quantitative analysis indicated most of the products continued low concentrations of the cyclic siloxanes (<0.6% for D4; <0.2% for D5, and < 1.7% for D6). In an exposure and risk assessment analysis of D4 in consumer products available in Denmark, D4 was identified in Pop it toys, teething rings for babies, iPad and tablet covers and watch straps; however, results indicated no migration of D4 above the limit of detection (LOD) was identified from any of the products (Danish Environmental Protection Agency, Poulsen et al. 2022).

The Norwegian Environmental Agency (NEA 2021; NILU 2020) conducted a survey of air emissions of volatile organic chemicals (VOCs) including D4, D5, and D6 from children’s toys and noted air concentrations of D4 and D5 in indoor breathing zones. Dudzina et al. (2014) analyzed cosmetic and personal care products (PCP) containing various concentrations of D4, D5, and D6 and concluded that concentrations of D5 were highest in all product categories and ranged from 0.01% in hand creams to more than 35% in deodorants. Dudzina et al. (2014) also noted that the lower concentrations of D4 in these products suggested D4 is primarily present as an impurity and might be the result of industry phasing out the use of D4. The use of D4, D5, and D6 in consumer products in Europe may soon change based upon recommendations from the European Chemicals Agency (ECHA 2021) to restrict the use of D4, D5, and D6 in cosmetics and consumer products that are intended to stay in prolonged contact with the skin or hair.

Because of their utilization in consumer products and potential for human dermal exposure, a more comprehensive understanding of dermal absorption and overall fate of siloxanes following dermal treatment is important. The purpose of this investigation was to identify and review available data associated with dermal absorption/penetration and fate of the most commonly used siloxane substances. This review focused on three cyclic siloxanes (D4, D5, and D6) and 5 linear siloxanes (L2, L3, L4, L5, and PDMS).

Materials and methods

In order to identify the relevant literature available regarding dermal absorption of cyclic and linear siloxanes, a comprehensive literature search was conducted. The initial step was development of a literature search strategy that included keywords associated with common chemical names and synonyms for the cyclic and linear siloxanes of concern. These keywords and synonyms were used in combination with keywords relating to the specific route of exposure (i.e., dermal) and keywords associated with the outcome of interest (i.e., absorption) of siloxanes as presented in Table 1.

Table 1. Literature search, keywords and number of articles identified.

Chemical name		Route		Outcome	Hits
octamethylcyclotetrasiloxane OR cyclotetrasiloxane OR D4	AND	dermal OR skin OR percutaneous	AND	absorption OR penetration OR uptake OR distribution OR transport OR metabolism OR metabolite OR excretion OR half-life	221
decamethylcyclopentasiloxane OR cyclopentasiloxane OR D5	AND	dermal OR skin OR percutaneous	AND	absorption OR penetration OR uptake OR distribution OR transport OR metabolism OR metabolite OR excretion OR half-life	149
dodecamethylcyclohexasiloxane OR cyclohexasiloxane OR D6	AND	dermal OR skin OR percutaneous	AND	absorption OR penetration OR uptake OR distribution OR transport OR metabolism OR metabolite OR excretion OR half-life	76
polydimethylsiloxane OR PDMS or dimethicone	AND	dermal OR skin OR percutaneous	AND	absorption OR penetration OR uptake OR distribution OR transport OR metabolism OR metabolite OR excretion OR half-life	243
Hexamethyldisiloxane OR L2	AND	None^1	AND	None^1	247
Octamethyltrisiloxane OR L3	AND	None^1	AND	None^1	25
Decamethyltetrasiloxane OR L4	AND	None^1	AND	None^1	15
Dodecamethylpentasiloxane OR L5	AND	None^1	AND	None^1	12
siloxane	AND	dermal OR skin OR percutaneous	AND	absorption OR penetration OR uptake OR distribution OR transport OR metabolism OR metabolite OR excretion OR half-life	459

¹No results were found when keywords other than the chemical name were added. These results reflect only searching for the chemical name.

Literature searching was conducted using the United States National Library of Medicine’s PubMed database, and searching was limited to literature published in English. The searches were conducted on April 28, 2022. As noted in Table 1, for L2, L3, L4, and L5, when the chemical names were combined with any other keywords, PubMed searches returned no results; therefore, these searches were conducted using the chemical names only. Literature searches were undertaken separately for each string of terms for each siloxane and number of articles identified for each search presented in Table 1. Each search was then combined into a reference management database, EndNote X7, and duplicates were removed resulting in a total of 258 articles. Following a screening of titles and abstracts focused on identifying both in vivo and in vitro studies for the cyclic and linear siloxanes of interest, 13 relevant articles were obtained. In addition, Dow Corning Corporation provided 14 unpublished study reports of in vitro and in vivo studies conducted with multiple siloxanes. Searches of authoritative websites and review articles focused on siloxanes were also reviewed in order to verify the literature searches identified all the relevant data. All identified data were critically reviewed and are organized by cyclic and linear siloxanes and summarized in the following sections.

Results

Cyclic siloxanes

The cyclic siloxanes, D4, D5, and D6 are employed in dermatological consumer products and cosmetics for both adults and children (Johnson et al. 2011). D4, D5, and D6 have all been identified by the Organization of Economic Co-operation (OECD) and the United States Environmental Protection Agency (USEPA) as high production volume (HPV) chemicals. Cyclic siloxanes are reliable candidates for use in consumer products, typically as emollients or solvents, due to their volatility, low surface tension, transparency, and hydrophobic nature (Biesterbos et al. 2015; Wang et al. 2009). All three cyclic siloxanes possess a specific affinity for the stratum corneum, more so for D6 compared to D4 and D5 (Krenczkowska et al. 2020); however, absorption through the skin is dependent upon the physical and chemical properties of D4, D5, and D6, which are summarized in Table 2 (Wang et al. 2009).

Table 2. Physical and chemical properties of D4, D5, and D6.

Cyclic Siloxane	Molecular Weight^a (g/mole)	Boiling Point^a (°C)	Density^a (kg/m^3)	Vapor Pressure^a (Pa) 25°C	Water Solubility^a (mg/L) 25°C	Log Kow^b	Air-water partition coefficient (Kaw)^b (Dimensionless)	Temperature at which the Kaw was measured^b (K)	Henry’s Law Constant^c (Pa m^3/mol)
D4	296.6	175	950	132	0.056	6.98	489.78	295.85	1,200,000
D5	370.8	211	954	23	0.017	8.07	1348.96	297.75	3,340,000
D6	444.9	245	963	3	0.005	8.87	1023.29	296.75	2,520,000

^aWang et al. (2009).

^bXu et al. (2014).

^cHenry’s Law constant can be estimated from the equation K_aw (dimensionless) = K_H/R x T, where K_aw is the chemical-specific air-water partition coefficient (dimensionless), KH is Henry’s Law Constant (Pa m³/mol), R is the universal gas constant of 8.314 Pa m³ mol⁻¹ K⁻¹ and T is the temperature that the K_aw was measured (K). Solving for K_H, the equation is K_H = K_aw x R x T (Charles and Destaillats 2005; Majer, Sedlbauer, and Bergin 2008).

D5 and to a lesser extent D6 are the predominant substances used in consumer products, whereas D4 is detected in smaller amounts (Biesterbos et al. 2015). Horii and Kannan (2008) reported that concentrations of cyclic siloxanes in consumer products ranged from < 0.35 to 9380 µg/g, <0.39 to 81,800 µg/g, and < 0.33 to 43,100 µg/g for D4, D5, and D6, respectively. An analysis of consumer products containing D4 available in Denmark included Pop in toys, teething rings for babies, iPad and tablet covers and watch straps (Danish Environmental Protection Agency, Poulsen et al. 2022). Migration analysis of these products indicated no migration of D4 above the LOD was identified from any of the products. The Norwegian Environmental Agency (NEA 2021; NILU 2020) conducted a survey of air emissions of VOCs from children’s toys including D4, D5, and D6 and noted increased air concentrations of D4 and D5 measured in the breathing zones at bedside. The utilization of PCP is the primary source of human dermal exposure to cyclic siloxanes (Biesterbos et al. 2015). Therefore, the Scientific Committee on Consumer Safety in Europe (SCCS 2010) and the Cosmetic Ingredient Review Expert Panel in the US (Johnson et al. 2011) assessed the health implications of the use of cyclic siloxanes in PCP and concluded that cyclic siloxanes are safe with regard to the present practices of use and their concentrations in PCP.

Octamethylcyclotetrasiloxane (D4)

Octamethylcyclotetrasiloxane or D4 (CAS Registry Number 556-67-2) is a low-molecular-weight volatile cyclic siloxane that serves predominately as a monomer or intermediate in the production of silicon-based polymers employed in industrial and consumer applications (Franzen et al. 2017). The low water solubility (0.056 mg/L) of D4, and high lipophilicity may increase the potential for dermal exposure to consumers, workers involved in the manufacture of D4, and workers involved in the production of polymers and products containing D4; however, the high volatility of D4 may limit dermal exposure due to volatilization from the skin surface (Franzen et al. 2017). The presence of D4 in 252 cosmetic and PCP was evaluated and D4 was found in 4.8% of the products tested (Wang et al. 2009). The levels of D4 in products ranged from 0.01 mg/g wet weight in hair spray to 3 mg/g wet weight in antiperspirant. Several published studies have investigated the dermal absorption characteristics of D4 in in vivo studies conducted in humans (Biesterbos et al. 2015; Reddy et al. 2007), in vivo studies conducted in rats (Dow Corning Corporation 2000b; Jovanovic et al. 2008), and ex vivo studies conducted with human skin (Dow Corning Corporation 1998; Jovanovic et al. 2008; Krenczkowska et al. 2019, 2020; Plotzke et al. 2000; Zareba et al. 2002).

As part of a global “harmonized” risk assessment, global exposure information combined with a Monte Carlo analysis was incorporated to determine the most significant routes of exposure for humans to D4 (Gentry et al. 2017). The Monte-Carlo-based probabilistic assessment for D4 included age-dependent (adult and children) and exposure-route-dependent consumer or general population scenarios (dermal, ingestion, and inhalation routes) with each product type examined independently by age and exposure scenario. The analysis included separate route-specific estimates made for male or female children 0–6 months, 6 months to 4 years, 4–11 years; teens 12–19 years; and adults 20–59 years and 60+ years. In addition, combined males and females for the ages of 0–6 months, 7–11 months and 1–2 years was stratified by breastfed versus non-breastfed. A non-gender-specific population, children ages 2–4 years was also considered. The results of the analysis indicated that in all cases, specific PCP use (body lotion, hair spray, foundation, after shave and antiperspirants) by adults provided the highest contribution to potential D4 exposure measured in grams per application for dermal exposure and µg per cubic meter in air for inhalation exposure. Similar estimates of intake also measured in grams per application for dermal exposure and µg per cubic meter in air for inhalation, were noted in children. The estimate of D4 intake from body lotion in 4- to 11-year-olds was the largest intake estimated from the Monte Carlo analysis and was within a factor of 1.5 of the mean estimates of intake for adult females in the 20–59-year-old group.

Investigations were conducted in human volunteers to estimate dermal absorption based upon the levels of D4 in expired air (Biesterbos et al. 2015; Dow Corning Corporation 2000c) and blood (Dow Corning Corporation 2000c) following dermal exposure. Biesterbos et al. (2015) conducted a series of experiments in which 15 volunteers were exposed to D4 as a pure substance or as an ingredient of a PCP. In the experiments, the 15 volunteers were exposed to 2.5 mg of radiolabeled D4 per cm² delivered as neat D4, a night cream containing D4 and D5, a deodorant containing D4 and D5, or a combination of the night cream and deodorant containing D4 and D5. The products were applied to forearm skin every 10 min for 60 min and then volunteers provided exhaled air samples at 65, 70, and 75 min. Prior to the beginning of the experiment, baseline levels of D4 and D5 were measured in all of the participants of the dermal exposure experiments and each volunteer completed a questionnaire and 24-hr PCP diary. In addition, prior to exposure, 6 of the volunteers served as controls and D4 and D5 were placed on an artificial arm next to each volunteer and the concentrations of D4 and D5 in end-exhaled air were measured. During the experiments, to prevent inhalation of D4 from ambient air after each exposure period, a fume hood was placed over the head of each volunteer. Every hr the participant was asked to provide a urine sample, during which time no attempt was made to prevent inhalation of ambient air.

The measurement of end-exhaled air has been reported be a reliable estimate of blood concentrations or internal dose, and because of their high vapor pressure and low blood:air partition coefficients, D4 and D5 are excellent candidates for detection in end-exhaled air. Biesterbos et al. (2015) reported no significant dermal uptake of neat D4 or D4 and D5 from the deodorant and night cream formulation containing both D4 and D5 based on end-exhaled air. The measured concentrations in exhaled air for the 6 control subjects ranged from 0.8 to 3.5 ng/L (D4) and from 0.8 to 4.0 ng/L (D5), which were considered background levels. Baseline levels in the 15 volunteers, measured prior to exposures ranged from 1.9 to 44.8 ng/L for D4 and 1.6 to 44.4 ng/L for D5. Following exposure, the D4 concentrations in the end-exhaled air of the exposed groups fluctuated during the exposure and post-exposure periods. The individual peak levels of D4 in expired air occurred during exposure and ranged from 7.5 to 280 ng/L. Biesterbos et al. (2015) concluded, based on measurements of D4 in end-exhaled air, that dermal absorption of D4 was minimal and the concentrations of D4 measured in end-exhaled air were similar to background levels observed following the nonuse of PCP for 24 hr.

Dow Corning Corporation (2000c, 2000) measured exhaled air, plasma and blood concentrations of D4 in 6 healthy volunteers following dermal application of approximately 1.4 g (males) or 1 g (female) of radiolabeled D4 to the axilla. Blood and exhaled air samples were collected prior to exposure and at 1, 2, 4, 6, and 24-hr following application. As noted in Table 3, the highest D4 levels detected in expired air, blood and plasma were reported at 1 hr following application in both men and women. D4 was detected in the expired air of all subjects at all time points with peak levels reported at 1 hr that ranged from 3.1 to 239.3 mg/L. D4 was also detected in the blood and plasma of all subjects at 1, 2, 4, and 6 hr after application with peak levels in the blood reported at 1 hr ranging from 0.57 to 5.67 ng/g for blood, and in the plasma ranging from 0.85 to 7.02 ng/g. In women, the mean blood peak level of D4 was 4.45 ng/gm (SD 1.10 ng/g) and for men it was 1.30 ng/gm (0.77 ng/g), with the differences in D4 blood and plasma levels between men and women being significantly different at 1-, 2-, and 4-hr following application. According to Plotzke et al. (2000a) plasma and blood levels of D4 were significantly correlated, while the exhaled air levels generally reflected blood and plasma levels but were not correlated.

Table 3. The average concentrations of D4 in exhaled air, blood and plasma in humans following dermal exposure¹.

		Females				Males
Biological media	Time (h)	1	2	3	Average	1	2	3	Average
Exhaled air (ng/L)	0^2	4.0	0.2	0.1	1.4 ± 2.2	0.1	0.2	ND	0.165
	1	26.7	239.3	66.9	110 ± 110	15.2	72.7	3.1	30 ± 37
	2	12	60.3	13.2	29 ± 28	5.5	11.5	1.5	6.2 ± 5.0
	4	8.7	14.3	5.9	9.6 ± 4.3	1.9	4.6	0.9	2.5 ± 1.9
	6	5.2	7.2	3.2	5.2 ± 2.0	1.1	3.3	0.8	1.7 ± 1.4
	24	1.1	1.3	0.5	0.97 ± 0.42	0.3	0.9	0.2	0.48 ± 0.37
Blood (ng/g)	0	ND	ND	ND	ND	ND	ND	ND	ND
	1	4.16	5.67	3.53	NR	0.57	2.10	1.22	NR
	2	3.35	4.06	1.95	NR	0.37	1.24	0.89	NR
	4	1.82	2.01	1.07	NR	0.26	0.91	0.58	NR
	6	1.43	1.11	0.54	NR	0.17	0.80	0.61	NR
	24	NQ	0.19	0.15	NR	ND	0.18	NQ	NR
Plasma (ng/g)	0	ND	ND	ND	ND	ND	ND	ND	ND
	1	6.5	7.02	3.94	5.8 ± 1.6	0.85	2.58	1.88	1.8 ± 0.9
	2	4.77	4.84	2.19	3.9 ± 1.5	0.58	1.77	1.36	1.2 ± 0.6
	4	2.72	2.34	1.24	2.1 ± 0.8	0.45	0.53	1.07	0.67 ± 0.34
	6	2.13	1.26	0.49	1.3	0.3	0.66	0.88	0.61 ± 0.29
	24	NQ	0.31	0.12	-^3	ND	0.24	0.30	-^3

¹Reported as the individual concentrations or average concertation ±1 Standard Deviation for three men or women. Individual concentrations reported in Dow Corning Corporation (2000c) and Reddy et al. (2007). Averages and standard deviations reported in Reddy et al. (2007).

²Samples at this time point had a small variable background radiolabel level due to the reuse of a Hans Rudolf nonrebreathing valve*, despite efforts to decontaminate the valve between uses.

³For one subject, the amount of D4 was below the detection limit or the limit of quantification, and so the average value was not reported.

NR = Average concentrations and standard deviation were not reported by Reddy et al. (2007).

ND = Not detected.

NQ = Not quantifiable.

* A specific valve with three separate ports for inhalation, exhalation and the mouth that prevent rebreathing once exhaled.

The dermal absorption data for D4 (Dow Corning Corporation 2000c; Plotzke et al. 2000) from human volunteers was then used by Reddy et al. (2007) to estimate pharmacokinetic parameters for a physiologically based pharmacokinetic (PBPK) model to describe the uptake of D4 from skin. The average concentrations of D4 in exhaled air and plasma in males and females are noted in Table 3 (Dow Corning Corporation 2000c; Reddy et al. 2007). The concentrations of D4 were higher in women compared to men; however, by the end of the exposure period, concentrations returned to background levels in both genders.

Due to the volatility of D4, the evaporation rate of D4 was experimentally determined to be 0.15 ± 0.04 mg/cm²/min. In addition to estimating the evaporation rate, the compartmental PBPK model included volatilization of the applied chemical from the skin surface, diffusion of absorbed chemical back to the skin surface, uptake from the skin compartment into blood, and a storage compartment within the skin.

The results from the dermal exposure study in male and female volunteers (Dow Corning Corporation (2000c) were then used to estimate model parameters for the human PBPK model for D4 developed previously (Reddy et al. 2003). A summary of the results of the PBPK model calculations are shown in Table 4 (Reddy et al. 2007). Based upon PBPK model estimates, only 0.3% and 0.12% of D4 reached systemic circulation in women and men, respectively. Model simulations suggested that the test articles may have been absorbed into the stratum corneal layer of the skin; however, more than 99% of what entered the skin diffused back to the skin surface and evaporated after exposure was terminated. Exhalation was estimated to be the major route of elimination of D4 and D5 from systemic circulation. For example, of the fraction of D4 estimated by the human PBPK model to be absorbed systemically in women (2.9 mg of approximately 1 g (female) of radiolabeled D4 applied), 83% was predicted to be exhaled and 10% metabolized within 24 hr (Table 4).

Table 4. Summary of PBPK model calculations for human D4 dermal Exposure¹.

Model calculation	Women	Men
Maximum concentration in plasma (µg/L)	105	54
Time that maximum concentration occurred (min)	10	10
Percent of applied dose that was absorbed systemically	0.30	0.12
Amount systemically absorbed (mg)	2.9	1.6
Amount eliminated by exhalation in 24 hr (mg)	2.4	1.4
Percent of systemically absorbed dose eliminated by exhalation in 24 hr	83	88
Amount metabolized in 24 hr (mg)	0.30	0.15
Percent of systemically absorbed dose metabolized in 24 h	10	9.4

¹The values reported here are all model simulated.

The dermal absorption of D4 was also assessed in an in vivo rat study (Dow Corning Corporation 2000b; Jovanovic et al. 2008), the results of which are comparable to those reported in human in vivo studies. In this study, female rats were assigned to one of three dose levels of 10, 4.8, or 2 mg/cm² (high, intermediate, or low, respectively) as presented in Table 5. Separate groups of female rats were cannulated via the jugular vein and used for a blood kinetics (BK) group. All animals received one dose of radiolabeled D4 applied using an aluminum skin depot with charcoal basket to collect any volatilized test article. A hole made in the plastic cap of the charcoal baskets allowed air to circulate providing semi-occluded conditions. An additional “wash group” that included 4 rats was added to evaluate the disposition of residual D4 following soap and water wash. Animals were housed in Roth-style glass metabolism cages, with rats in the wash group removed from cages after 24 hr, dose sites washed, charcoal baskets replaced, and the animals returned to cages for an additional 144 hr. Urine, feces, and expired air were collected in the metabolism cages. At the end of each exposure period, or at 168 hr for the wash group, blood was collected via cardiac puncture, charcoal baskets removed and extracted. In addition, the skin was washed, tape stripped, excised, and solubilized along with the remaining carcasses. The radioactivity content was measured by liquid scintillation counting, and % dose absorbed determined as the amount of radioactivity in expired volatiles, carcasses, excreta, skin, and cage rinses.

Table 5. Dosing schedules in female rats exposed dermally to D4.

Group Assignment	Dose (mg/cm^2)	Exposure Period (hours)	Number of rats
Mass Balance Control	0	24	6
Mass Balance High Dose	10	1	4
		6	4
		24	4
		168 (24 + 144 h after skin was washed)	4
Mass Balance Mid Dose	4.8	1	4
		6	4
		24	4
		168 (24 + 144 hr after skin was washed)	4
Mass Balance Low Dose	2	1	4
		6	4
		24	4
		168 (24 + 144 hr after skin was washed)	4
Blood Kinetics	0	10	2
	10	10	6

Results indicated that most of the test article evaporated from the skin surface within 6 hr of exposure (Dow Corning Corporation 2000b; Jovanovic et al. 2008) (Table 6). The dose recovered from the skin surface was less than 1% at all dose levels and all time points except at the high dose after 1 hr treatment where cumulative radioactivity content in expired volatile traps increased from 26.93 µg to 59.18 µg between the 1^st and 2^nd hr exposure. The average % applied dose absorbed and recovered from the excreta, excised exposure site, expired air and carcass was between 0.6% and 0.95% for all dose levels and all exposure times. There was a significant decrease in % dose absorbed over time, with absorption after 24 hr significantly lower than absorption at 1 hr (absorption at 6 hr was not markedly different from 1 to 24 hr). The % dose that remained in the skin over 24 hr, fell from 0.82% to 0.22% at the high dose, 0.52% to 0.23% at the intermediate dose, and 0.47% to 0.21% at the low dose.

Table 6. Disposition of absorbed radioactivity following dermal application of neat¹⁴C-D4 to female Fischer 344 Rats¹.

Dose (mg/cm^2)	Time (hr)	Expired Volatiles	CO2	Excised Skin Application Site	Urine	Feces + Cage Rinse	Carcass	Total Absorbed^2	% of absorbed dose recovered in skin
Mass Balance High Dose
10	1	0.11 ± 0.02	0.00 ± 0.00	0.82 ± 0.01*	0.00 ± 0.00	0.00 ± 0.00	0.02 ± 0.01	0.95 ± 0.12*	86.3
	6	0.39 ± 0.06	0.00 ± 0.00	0.34 ± 0.04	0.00 ± 0.00	0.00 ± 0.00	0.05 ± 0.01	0.79 ± 0.09	43.
	24	0.28 ± 0.05	0.01 ± 0.00	0.22 ± 0.03	0.03 ± 0.01	0.01 ± 0.00	0.05 ± 0.01	0.61 ± 0.09	36.1
	168	0.12 ± 0.01	0.03 ± 0.00	0.08 ± 0.10**	0.08 ± 0.01	0.02 ± 0.01	0.02 ± 0.00	0.35 ± 0.01**	22.9
Mass Balance Intermediate Dose
4.8	1	0.27 ± 0.09	0.00 ± 0.00	0.52 ± 0.05*	0.00 ± 0.00	0.00 ± 0.00	0.02 ± 0.01	0.80 ± 0.06*	65.
	6	0.26 ± 0.07	0.00 ± 0.00	0.38 ± 0.03	0.00 ± 0.00	0.00 ± 0.00	0.04 ± 0.00	0.69 ± 0.05	55.1
	24	0.24 ± 0.05	0.01 ± 0.00	0.23 ± 0.00	0.04 ± 0.00	0.01 ± 0.00	0.04 ± 0.00	0.57 ± 0.05	39.7
	168	0.28 ± 0.01	0.01 ± 0.00	0.09 ± 0.01**	0.07 ± 0.01	0.01 ± 0.00	0.02 ± 0.00	0.47 ± 0.01**	19.1
Mass Balance Low Dose
2	1	0.30 ± 0.05	0.00 ± 0.00	0.47 ± 0.01*	0.00 ± 0.00	0.00 ± 0.00	0.02 ± 0.00	0.79 ± 0.06*	59.5
	6	0.42 ± 0.13	0.00 ± 0.00	0.32 ± 0.02	0.00 ± 0.00	0.00 ± 0.00	0.03 ± 0.01	0.78 ± 0.16	41.
	24	0.47 ± 0.09	0.01 ± 0.00	0.21 ± 0.01	0.03 ± 0.00	0.00 ± 0.00	0.02 ± 0.00	0.76 ± 0.08	27.6
	168	0.32 ± 0.08	0.02 ± 0.00	0.09 ± 0.01**	0.06 ± 0.01	0.01 ± 0.00	0.01 ± 0.00	0.51 ± 0.08**	17.6

¹Average % applied dose ± standard error.

²Total dose absorbed includes the sum of expired volatiles, CO₂, excised skin at the application site, urine, feces and cage rinse, and carcass.

*Significant difference (p = 0.05) in % dose absorbed between 1–24 hr at all dose levels.

**Significant difference (p = 0.05) in % dose absorbed between 24–168 hr at all dose levels.

Results from the female rats in the blood kinetics groups administered 10 mg/cm² D4 indicated concentrations of D4 in all blood samples were not markedly different from controls. As predicted by the PBPK model simulations of Reddy et al. (2007), data indicated that D4 partitioned to the upper layer of the skin over time and evaporated from the skin surface (Dow Corning Corporation 2000b; Jovanovic et al. 2008). No significant differences in % dose absorbed was detected among dose groups at any time indicating the absorption of D4 did not appear to be dose dependent. In the wash group at 168 hr post exposure, the total dose recovery at all three dose levels was higher than 91%. Absorption in the wash group was 0.35, 0.47, and 0.51% at the high, intermediate, and low doses, respectively, which was significantly lower than absorption at 24 hr treatment (0.61, 0.57, and 0.76% in the high, intermediate, and low-dose groups, respectively).

Utilizing multiple in vitro and/or ex vivo percutaneous absorptions models, several investigators conducted human skin studies to assess dermal absorption of D4 (Jovanovic et al. 2008; Krenczkowska et al. 2019, 2020; Mojsiewicz-Pieńkowska et al. 2020; Zareba et al. 2002). Due to the differences in the permeability of human and animal skin, a study using human skin/nude mouse model was conducted to determine percutaneous absorption of neat D4 (Dow Corning Corporation 1999; Zareba et al. 2002). Human fetal forearm skin, obtained from aborted fetuses at gestation weeks 16–22, was grafted subcutaneously onto female BALB/C nude mice. The animals were kept for 2–4 months after the transplant to enable the graft to heal and the skin to fully develop barrier properties. A dose of radiolabeled D4 (15.7 mg/cm²) was applied to skin depots fixed to the human skin grafts on 4 female mice for 24 hr. After the exposure duration, the skin depots were removed, skin areas were cleaned, and mice observed for an additional 72 hr for urine, feces, and potassium hydroxide (KOH) trap samples. To evaluate the skin distribution of D4, unlabeled D4 (15.7 mg/cm²) was applied to skin depots fixed to the human skin grafts of 4 female mice. A charcoal basket was immediately placed into the skin depot casing and animals placed into metabolism cages. A hole was placed in the cap of the charcoal basket to provide semi-occluded conditions.

After 24 hr treatment almost all of the applied test substance evaporated from the application site (94.59% of the applied dose) and detected in the charcoal baskets (Dow Corning Corporation 1999; Zareba et al. 2002) (Table 7). The application site was found to contain 0.01% of the applied dose and mean recovery was 96.35 ± 12.28%. The maximum absorption for the applied dose of D4 was estimated to be 1.09% based upon the sum of the radioactivity of all measurements and assuming no leakage occurred from the application cell. Elimination via expired air consisted of 42% of what was absorbed and 49% of the absorbed dose of D4 was excreted in urine and feces. Results of the skin distribution study showed total concentrations of D4 in the human epidermis, dermis, and adipose tissue after 24 hr were 470, 220, and 75 ng of D4, respectively (Dow Corning Corporation 1999; Zareba et al. 2002). The mean distribution of total D4 was 61%, 29%, and 10% in the skin, epidermis, and adipose tissue, respectively. Therefore, most of the applied dose (61%) remained in the upper skin layer or the epidermis.

Table 7. Disposition of radioactivity in Epidermis, dermis, and adipose tissue of human skin following application of neat D4 to human skin/Nude mouse model.

Tissue	Total ng D4				Mean ± SD N = 4	% applied dose
	Animal 1	Animal 2	Animal 3	Animal 4
Epidermis	846.00	244.50	533.60	254.12	470 ± 285	61
Dermis	308.00	52.00	375.45	145.85	220 ± 148	29
Adipose	121.80	23.40	0.00	156.65	75 ± 76	10

Jovanovic et al. (2008) noted similar results when human abdominal skin samples from 6 cadavers were exposed in vitro using flow-through chambers to radiolabeled D4 either as an antiperspirant formulation or neat (Dow Corning Corporation 1998; Jovanovic et al. 2008). The neat dosing solution was a mixture of radiolabeled-D4 and unlabeled D4, while the formulated test substance was a mixture of the neat solution and a generic antiperspirant formulation. Skin samples were dermatomed to a thickness ranging from 300 to 500 µm to separate the epidermis from dermis and then D4 was applied for 24 hr at a mean dose of 10.7 mg/cm² (neat D4) or 9.5 mg/cm²(formulated D4). Immediately after application, charcoal baskets were placed above the skin and secured into a custom designed cap to capture any volatilized material. After 24 hr, the charcoal baskets were removed and extracted, skin washed and solubilized, and receptor fluid collected. Data indicated that 0.5% ± 0.07% of the applied dose of neat D4 and 0.49% ± 0.19% of the formulated D4 was absorbed, and the difference in absorption between neat and formulated D4 was not significant (Table 8). The % applied dose recovered was 91.6% ± 3.4% and 103.8% ± 1.9% for neat D4 and formulated D4, respectively. The majority of the test article volatilized and was collected into charcoal baskets. The total amount of D4 absorbed was approximately 0.5% for both formulations and less than 4% remained on the skin surface.

Table 8. In vitro disposition of radioactivity 24 hr after application of¹⁴C-D4 to human skin¹.

Formulation	N^2	Volatilized	Skin Surface^3	Skin	Receptor Fluid	Total Absorbed^4	Recovery	Steady State Flux (µg/cm^2/hr)	Kp (cm/hr)
Neat	6	88.2 ± 3.4	2.96 ± 0.1	0.49 ± 0.07	0.01 ± 0.00	0.50 ± 0.07	91.6 ± 3.4	0.06^5	6.2×10^−8
Antiperspirant	4	99.3 ± 1.8	3.98 ± 0.19	0.48 ± 0.18	0.01 ± 0.00	0.49 ± 0.19	103.8 ± 1.9	0.04^6	6.3×10^−8

¹% applied dose ± Standard error.

²Number of individual donor skin specimens that showed acceptable skin integrity. Skin specimen from each donor was tested in duplicates.

³Total radioactivity found in extracts of swabs, tape strips, and diffusion cells.

⁴Total radioactivity associated with the skin and receptor fluid.

⁵Between 5–15 hr.

⁶Between 6–21 hr.

The steady state flux for neat D4 was 0.06 µg/cm²/hr; while the steady state flux for formulated D4 was 0.04 µg/cm²/hr (Dow Corning Corporation 1998; Jovanovic et al. 2008). The lag time for both neat and formulated D4 was approximately 3 hr and permeability constants (Kp), calculated by dividing the steady state flux by the concentration of D4 in the dosing solution, were 6.2 × 10⁻⁸ cm/hr for neat D4 and 6.3 × 10⁻⁸ cm/hr for formulated D4. Overall, Jovanovic et al (2008) and Dow Corning Corporation (1998) concluded that the results of the study demonstrated the majority of the D4, 88.17 ± 3.38% for neat and 99.29 ± 1.8% for the formulated, was volatilized from the skin and % absorbed dose in the skin was 0.47 ± 0.07% and 0.45 ± 0.18% for neat and formulated D4, respectively.

A similar study was conducted in swine skin exposed in vitro using the Bronaugh flow-through diffusion cell system to three personal care application formulations of radiolabeled D4 (skin moisturizer, roll-on antiperspirant or cuticle coat) (Dow Corning Corporation 2006a). The swine skin samples were dermatomed to obtain a split thickness of 320–500 µm. Each D4 formulation was prepared in two different D4 concentrations as presented in Table 9.

Table 9. D4 preparations and concentrations.

Experiment	Dose Formulation/Nominal Concentration	D4 Content (%)
1	Skin moisturizer/5%	4.97 ± 0.3
2	Skin moisturizer/40%	41.7 ± 1.0
3	Antiperspirant/10%	10.6 ± 0.1
4	Antiperspirant/60%	62.2 ± 2.6
5	Cuticle coat/50%	51.6 ± 0.9
6	Cuticle coat/90%	95.8 ± 2.2

The study included six experiments, in which three skin samples were evaluated each day. The skin samples were placed in diffusion cells in six replicates and the radiolabeled formulation applied at a targeted dose of 10 mg/cm² of skin (Dow Corning Corporation 2006a). The skin was exposed to the formulation for a minimum of 24 hr. Receptor fluid was collected at every hr for the first 6 hr and then every third hr up to 24 hr (experiments 1 and 2) or every hr for the first 6 hr and then every other hr up to 24 hr (Experiments 3–6). Immediately after application a charcoal basket was placed in each diffusion cell to capture any volatilized material. Data demonstrated that, as noted in human skin samples, regardless of the formulation, the majority of applied D4 volatilized from the swine skin surface (>99.5% of the total recovered dose from all formulations) and was captured in the charcoal baskets. The total % dose absorbed in swine skin and receptor fluid was < 0.05% of applied dose in all experiments. A small amount of D4 (≤0.01%) penetrated through the skin with a mean cumulative penetration of D4 of less than 0.6 µg equivalents D4/cm². The estimated permeability coefficient (K_p) ranged from 1 × 10⁻⁷ cm/hr for the 5% skin moisturizer to 1.9 × 10⁻⁹ cm/hr for the 10% antiperspirant.

In vitro studies were also conducted that focused on the penetration of D4 through the stratum corneum, permation into the epidermis, dermis, and receptor fluid, as well as the diffusion pathway f D4 through human skin using Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectroscopy and gas chromatography-flame ionization detection (GC-FID) (Krenczkowska et al. 2019, 2020). These studies were conducted ex vivo using human abdominal skin samples extracted from 5 to 7 male and female cadavers within 24–48 hr following death. A static Franz type diffusion cell system was used to facilitate a 24-hr application of D4 to skin samples (stratum corneum side up). Krenczkowska et al. 2019; 2020) indicated that the D4 application amount of 100 µl was selected to correspond with an infinite dose of approximately 95,600 µg.

Based upon the ATR-FTIR results, D4 penetrated the stratum corneum and permeated the epidermis and dermis following the 24-hr exposure. Skin protein interactions were slightly altered from exposure to D4 (Krenczkowska et al. 2019). GC-FID results indicated D4 accumulated primarily in the stratum corneum (Krenczkowska et al. 2019, 2020). The cumulative dose of D4 identified in each skin layer is reported in Table 10. Permeation into the receptor fluid was also observed and suggested potential systemic absorption. Results of both studies confirmed that the highest amount of D4 following in vitro exposure to human skin may be found in the stratum corneum, with lesser amounts noted in the epidermis, dermis, and receptor fluid. In addition, microscopic images obtained through fluorescence microscopy indicated D4 disrupted the structure of the stratum corneum. D4 skin permeation occurred by trans-epidermal transport through the lipid matrix with disruption of corneocytes, as well as lipophilic canyons, indicating D4 might potentially alter the skin barrier.

Table 10. Cumulative dose of D4 in particular layers and fluid acceptor.

Sample (n = 5)	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2019)	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2020)
Stratum corneum	28.60	27.5
Epidermis	6.36	6.9
Dermis	5.12	5.1
Fluid Acceptor	2.41	4.1
Total Cumulative Dose	42.50	43.6

Mojsiewicz-Pieńkowska et al. (2020) used digital holographic microscopy (DHM), a skin-imaging technique to investigate the effects of D4 on the skin barrier. The stratum corneum is responsible for protection against various factors entering the skin and as reported by Krenczkowska et al. 2019;2020) D4 was reported to disrupt the structure of the stratum corneum. Mojsiewicz-Pieńkowska et al. (2020) used human cadaver skin prepared employing the same methods as used by Krenczkowska et al. 2019; 2020). D4 was directly applied (the exact dose of D4 applied was not reported by the study authors) on the epidermis surface in an amount that enabled coverage of the whole surface (approximately 1 cm²) and then the samples were incubated in a closed petri dish for 1 hr. Irreversible damage of the stratum corneum was reported following exposure to D4 which consisted of (1) destabilization of intercellular lipid lamellae and corneocyte structure, (2) loss of stratum cornea integrity and format of lacunae, and (3) changes in the surface geometrical topography of the stratum corneum. These results are similar to findings with other lipophilic solvents (Barba et al. 2016; Goffin, Letawe, and Piérard 1997).

To summarize, the results of the in vivo studies conducted in rodents and humans indicated that approximately 1% of the dermally applied dose of D4 is absorbed through the skin, with ≤ 0.5% of the applied dose of D4 passing through the skin to systemic compartments (Jovanovic et al. 2008; Reddy et al. 2007; Zareba et al. 2002). The majority of D4 applied to the skin volatilizes from the skin surface before dermal absorption occurs (Biesterbos et al. 2015; Jovanovic et al. 2008; Zareba et al. 2002). Studies performed in vitro support these findings with only minimal absorption of D4 occurring (0.5 to 1%) in human skin and < 0.05% in swine skin (Jovanovic et al. 2008; Zareba et al. 2002). Jovanovic et al. (2008) also reported that the largest amount of the absorbed dose (>90%) for the rat was found in the skin and had not passed through the skin layers and into the receptor fluid. The estimated dermal absorption of D4 and D5 in human volunteers from a PBPK model evaluation (Reddy et al. 2007) of blood time-course data (Plotzke et al. 2000) are also similar to findings of other in vivo and in vitro studies, with only 0.3% and 0.12% of D4 predicted to reach systemic circulation in women and men, respectively. Model simulations also estimated that while D4 may have been absorbed into the skin compartment, almost all of (99%) of what was retained in the skin moved back to the skin surface and evaporated after exposure. Studies using ATR-FTIR spectroscopy and GC-FID to determine the distribution of D4 at each skin layer noted that, for continuous exposure to D4 under conditions where evaporation was prevented, > 60% of the dermally applied D4 dose remained on the skin surface (Krenczkowska et al. 2019, 2020), while most of the applied dose that remained in the skin was found in the epidermis (61%), with only 29% detected in the dermis and 10% in the adipose tissue (Zareba et al. 2002). Irreversible damage of the stratum corneum was found following in vitro exposure to D4 which changed the surface geometrical topography of the stratum corneum possibly highlighting a mechanism by which D4 penetrates the skin barrier., similar to other lipophilic solvents (Barba et al. 2016).

Decamethylcyclopentasiloxane (D5)

Decamethylcyclopentasiloxane or D5 (CAS Registry Number 541-02-6) is used in a variety of consumer products, as well as an industrial intermediate in the production of polydimethylsiloxanes (Dekant and Klaunig 2016). Compared to D4, D5 is slightly less volatile and less soluble in water (Dekant and Klaunig 2016; Wang et al. 2009). Wang et al. (2009) reported that D5 was found in 14.3% of 252 products tested or almost 3-fold higher the number of products compared to D4, with concentrations of D5 in the products tested ranging from 0.02 mg/g wet weight in hairspray and deodorant to 683 mg/g wet weight in antiperspirant.

A global “harmonized” risk assessment including global exposure information combined with a Monte Carlo analysis conducted for D4 (Gentry et al. 2017) was conducted for D5 (Franzen et al. 2017) to determine the most significant routes of exposure for D5. The results of the analysis for D5 were similar to D4 and indicated that in all cases, consumer product use (body lotion, hair spray, foundation, after shave and antiperspirants) by adults provided the highest contribution to potential D5 exposure. In children, similar estimates of intake were noted. Diaper cream provided the highest contribution to potential D5 exposure for children under 4 years of age and body lotion for children 4–11 years old. Comparison of the values indicated that exposure of children was no more than 2-fold higher than that of adults.

The dermal absorption of D5 following in vivo exposure in humans (Biesterbos et al. 2015; Reddy et al. 2007) and rats (Dow Corning corporation 1996b; Jovanovic et al. 2008) and in vitro (Dow Corning Corporation 1996b; Dudzina et al. 2015; Jovanovic et al. 2008; Krenczkowska et al. 2019, 2020) has been investigated. Biesterbos et al. (2015) measured the concentrations of D5 in end-exhaled air, considered a reliable estimate of blood concentrations, following dermal exposure to neat D4 or deodorant or night cream formulations containing both D4 and D5. Biesterbos et al. (2015) exposed the forearms of 15 volunteers to neat D5, a night cream containing D4 and D5, a deodorant containing D4 and D5, or a combination of the night cream and deodorant at a dose of 2.5 mg of D5 per cm². To prevent the inhalation of volatilized D5 from the surface of the exposed skin, volunteers were seated with their exposed forearms inside a flow cabinet. Exposure was stopped after 1 hour and the product was washed from the skin surface and a fume hood was placed over the head of each volunteer to prevent inhalation of D4 from ambient air. Prior to the beginning of the experiment, baseline levels of D4 and D5 were measured in all of the participants of the dermal exposure experiments and each volunteer completed a questionnaire and 24-hrour PCP diary. In addition, prior to exposure, six of the volunteers served as controls and D4 and D5 were placed on an artificial arm next to each volunteer and concentrations of D4 and D5 in end-exhaled air measured. During the experiments, to prevent inhalation of D4 from ambient air after each exposure period, a fume hood was placed over the head of each volunteer. Every hr the participant was asked to take a bathroom visit to provide a urine sample, during which time no attempt was made to prevent inhalation of ambient air. Biesterbos et al. (2015) found no significant dermal uptake of neat D5 or D4 or any of the formulations containing both D4 and D5. Background levels of D5, based upon end-exhaled air concentrations from 6 control volunteers, ranged from 0.8 to 3.5 ng/L (D4) and from 0.8 to 4 ng/L (D5). Baseline levels in the 15 volunteers, measured prior to exposures ranged from 1.9 to 44.8 ng/L for D4 and 1.6 to 44.4 ng/L for D5. Following exposure, the D4 concentrations in the end-exhaled air of the exposed groups fluctuated during the exposure and post-exposure periods. The maximum concentration of D5 in end-exhaled air ranged from 3.8 to 605 ng/L and occurred during the exposure, as well as post exposure. Post-exposure peaks were assumed by Biesterbos et al. (2015) to be related to bathroom breaks without a fume hood. Evidence indicated that dermal absorption of D5, estimated based on analysis of D5 in end-exhaled air, was minimal and concentrations of D5 measured in end-exhaled air were similar to background levels observed following the nonuse of PCP for 24 hr.

Dow Corning Corporation (2000c, 2002) determined the dermal absorption of D5 applied to axilla of 3 men and 3 women for up to 24 hr. After the exposure period, the concentrations of D5 in the plasma and exhaled air were measured. The mean concentrations of D5 in exhaled air and plasma in males and females are presented in Table 11. The concentrations of D5 in exhaled air were higher in women compared to men for the first 2 hr following exposure. By the end of the treatment period, concentrations had not returned to background levels. Data from this study were later used to estimate fraction absorbed using a dermal exposure PBPK model (Reddy et al. 2007). Reddy et al. (2007) also determined that the evaporation rate of D5 at room temperature was 0.07 ± 0.03/cm²/min.

Table 11. Average concentrations of D5 in exhaled air and plasma in humans following dermal exposure¹.

		Females				Males
Biological media	Time (hr)	1	2	3	Average	1	2	3	Average
Exhaled air (ng/L)	0^2	2.2	3.0	2.3	2.5 ± 0.5	NQ	NQ	1.9	0.98 ± 0.82
	0.25	320	710	350	460 ± 220	700	150	1000	620 ± 430
	0.5	260	700	91	350 ± 310	1700	140	1000	930 ± 770
	0.75	420	370	260	350 ± 80	1600	680	210	830 ± 700
	1	120	360	150	210 ± 130	2300	220	230	920 ± 1200
	1.25	150	320	91	190 ± 120	520	280	180	330 ± 170
	1.5	220	180	47	150 ± 90	740	360	270	460 ± 250
	1.75	170	120	29	110 ± 70	670	300	98	350 ± 290
	2	350	110	36	160 ± 160	620	250	58	310 ± 280
	4	96	49	16	53 ± 40	120	24	22	55 ± 55
	6	64	15	7.5	29 ± 31	26	17	7.4	17 ± 9
	24	18	5.6	5.0	9.6 ± 4.5	3.9	13	6.3	7.6 ± 4.4
Plasma (µg/L)	0	ND	ND	ND	ND	ND	ND	ND	ND
	0.5	0.53	1.4	0.65	0.87 ± 0.49	1.2	0.25	1.1	0.86 ± 0.53
	1	0.95	2.0	0.77	1.2 ± 0.7	1.7	0.70	1.2	1.2 ± 0.5
	2	0.92	1.8	0.50	1.1 ± 0.7	1.6	0.97	0.90	1.2 ± 0.4
	4	0.90	1.5	0.34	0.91 ± 0.58	1.2	0.61	0.62	0.79 ± 0.31
	6	0.75	0.82	0.21	0.52 ± 0.33	0.88	0.51	0.47	0.49 ± 0.23

¹Reported as the individual concentrations or average concertation ±1 Standard Deviation for three men or women. Data reported in Reddy et al. (2007).

²Samples at this time point had a small variable background radiolabel level due to the reuse of a Hans Rudolf nonrebreathing valve, despite efforts to decontaminate the valve between uses.

ND = Not detected.

NQ = Not quantifiable.

Using a dermal model based on the human PBPK model for D5 developed by Reddy et al. (2003), together with the measured evaporation rate, model simulations of the dermal absorption of D5 estimated that 99% of what was absorbed into the skin moved back to the skin surface and was evaporated following treatment (Reddy et al. 2007). Very low systemic absorption was predicted: model calculations estimated only 0.05% (for women) and 0.05% (for men) of the D5 applied reached systemic absorption. Exhalation was estimated to be the major route of elimination. For example, of the fraction of D5 estimated by the human PBPK model to be absorbed systemically in women (0.48 mg of approximately 1 g (female) of radiolabeled D5 applied), 92% was predicted to be exhaled and 1.6% metabolized within 24 hr (Table 12).

Table 12. Summary of PBPK model calculations for human D5 dermal Exposure¹.

Model calculation	Women	Men
Maximum concentration in plasma (µg/L)	1.0	1.2
Time that maximum concentration occurred (min)	59	21
Percent of applied dose that was absorbed systemically	0.051	0.050
Amount systemically absorbed (mg)	0.48	0.67
Amount eliminated by exhalation in 24 hr (mg)	0.44	0.62
Percent of systemically absorbed dose eliminated by exhalation in 24 hr	92	93
Amount metabolized in 24 hr (mg)	0.0076	0.0097
Percent of systemically absorbed dose metabolized in 24 hr	1.6	1.4

¹The values reported here are all model simulated.

Jovanovic et al (2008, 2003b). evaluated the dermal absorption of D5. Female rats were dermally exposed to 10 mg D5/cm² for 6, 24, or 168 hr (24 hr exposure and 144 hr after washing of skin surface) (Groups 2, 4, and 5). Two female rats served as controls (Group 1) and an additional group of 4 euthanized rats (no expired air) exposed to 10 mg D5/cm² for 24 hr was also included (Group 3). The highest dose from the D4 study (Dow Corning Corporation 2000b; Jovanovic et al. 2008) was the only dose used based upon lack of absorption in the lower dose groups in the D4 study. Radioactivity measurements in urine, feces, CO₂, and expired or escaped volatiles (charcoal tubes) were taken at 6, 24, and 168 hr (wash group only). Only 0.27% of the applied dose of D5 was absorbed into and through the viable layers of skin following a 6-hrexposure, and less than 0.1% following 24 hr treatment (Table 13). The majority of D5 was volatilized and recovered from the skin surface. Over the 24-hr exposure period, less than 0.5% of the applied dose reached systemic circulation across all dose groups without any significant changes in the total dose absorbed. Total absorbed D5 was lower than that reported for D4 (Table 6); therefore, blood kinetics were not evaluated.

Table 13. Overall mass balance of radioactivity in female fischer 344 rats following dermal exposure of neat ¹⁴C-D5 (average percent of applied dose ± SE).

Group No.	Exposure Duration (hr)	Volatilized	Escaped Volatiles	Skin Surface	Excised Skin	Urine	Feces + Cage Rise	Carcass	Recovery	Absorbed	Systemically Absorbed^2	% absorbed dose recovered from skin
2	6	58.8 ± 2.3	0.20 ± 0.04	23.73 ± 1.66	5.55 ± 0.51	0.00 ± 0.00	0.00 ± 0.00	0.27 ± 0.12	88.5 ± 0.6	5.82 ± 0.49	0.27 ± 0.12	95.3
3^1	24	86.4 ± 4.8	7.75 ± 5.57	0.39 ± 0.22	0.25 ± 0.06	NA^5	NA	0.06 ± 0.03	94.9 ± 1.6	0.31 ± 0.06	NA	NA
4	24	80.4 ± 9.7	8.76 ± 8.41	0.17 ± 0.10	0.17 ± 0.02	0.01 ± 0.01	0.01 ± 0.00	0.06 ± 0.03	89.6 ± 2.1	0.24 ± 0.03	0.07 ± 0.04	71.9
5	168	88.8 ± 6.7	7.04 ± 4.17	0.05 ± 0.01	0.03 ± 0.01*	0.02 ± 0.01	0.01 ± 0.01	0.02 ± 0.01	95.9 ± 5.5	0.09 ± 0.03	0.06 ± 0.02	37.2

*Indicates a significant difference (p = 0.05) in % dose absorbed between 24-hr and 168-hr wash group.

¹Animals were euthanized immediately prior to dosing.

²Based upon radioactivity in urine, feces, cage rinses and carcass.

The animals in the wash group (Group 5) were used to determine the biological fate of D5 after percutaneous exposure (Dow Corning Corporation 2003b; Jovanovic et al. 2008). After 24 hr, any test article was washed from the skin and animals were monitored for the next 6 days (144 hr). At the end of the monitoring period, the mean total dose recovered was 95.93% of the applied dose. In the wash group (Group 5), the absorbed dose and % D5 that remained on the skin was significantly lower than findings from Group 4, which according to these investigators indicated the absorbed D5 diffused back to the skin surface and continued to evaporate over the additional 6 days. Overall, Jovanovic et al (2008) and Dow Corning Corporation (2003b) concluded that D5 was not significantly absorbed into systemic circulation.

Dow Corning Corporation (1996b) conducted an in vivo dermal absorption study in rats in order to examine the ability of in vitro techniques to predict in vivo dermal absorption. The in vitro study (Dow Corning Corporation 1996a), which ran concurrently, is also discussed below. For the in vivo dermal absorption study, 10 female and 12 male rats were divided into two test groups, an excreta group and a blood/tissue collection group. Approximately 20 µCi of radiolabeled D5 was delivered to each animal as 20 µl that were applied to 2.54 cm² of skin. The dose site was covered and radiolabeled D5 remained on the skin for 24 hr, after which the skin was cleaned, and then rats were returned to cages for further sample collection. At scheduled intervals, urine and feces (6, 12, 24, 48, 72 or 96 hr), blood (1, 2, 4, 6, 12, 24, 48, 72 or96 hr), and expired air (1, 2, 4, 6, 9, 12, 24, 48, 72 or 96 hr) were collected. Following sacrifice at 96 hr post exposure, the skin at the application site was stripped with tape to determine the amount of radioactivity that remained in the stratum corneum and then extracted. In the excreta group, 85% of total dose was volatilized from the skin surface and 0.35% was measured at the dose site. Less than 1% of the D5 applied to the skin was recovered in urine and carcass and trace levels were detected feces, expired air (CO₂ traps), and tissues. The total average amount of D5 absorbed through the skin was 0.8 ± 0.26% (Table 14). Activity in the blood was not detected at any time point sampled. The overall disposition of radioactivity was similar in the blood/tissue and excreta groups, with the liver and fat tissues found to contain the highest radioactivity concentrations and only trace levels present throughout the other sampled tissues. Due to the volatility of the test substance, total recovery values were not attempted, and expired air samples were not included.

Table 14. In vivo absorption of ¹⁴C-D5 following dermal exposure (% applied dose).

Study/Gender^1	Charcoal Basket^2	Charcoal Tubes^2	Skin Surface^3	Dose Site	Carcass	CO2	Urine	Feces	Biological samples^4	Total Absorbed^5	Total Recovery
In vivo Female	33.7 ± 4.1	47.62 ± 7.0	1.52 ± 1.31	0.03 ± 0.01	0.19 ± 0.07	0.04 ± 0.02	0.36 ± 0.17	0.19 ± 0.05	0.77 ± 0.30	0.81 ± 0.31	83.6 ± 4.4
In vivo Male	55.8 ± 5.6	32.6 ± 6.3	3.57 ± 0.99	0.59 ± 0.81	0.01 ± 0.01	0.04 ± 0.01	0.23 ± 0.04	0.04 ± 0.02	0.28 ± 0.05	0.87 ± 0.83	92.8 ± 2.6
Average Females-Males	ND	ND	ND	0.35 ± 0.64	ND	ND	ND	ND	0.45 ± 0.27	0.80 ± 0.62	ND

¹Data is expressed as the mean of 4 female or 6 male animals ± SD.

²Indicates leakage from the dose site.

³The combined values for the dosing chambers, dosing bandages, filter papers, soap gauze, final bandages, tape, Q-tip swab, and ethanol gauze.

⁴Total radioactivity found in Carcass, CO₂, Urine, and Feces.

⁵Total radioactivity found in Carcass, CO₂, Urine, Feces, and Dose Site.

⁶ND – No data. Values were not calculated.

In conjunction with the in vivo dermal absorption study in rats (Dow Corning Corporation 1996b), an in vitro study was conducted to evaluate the ability of in vitro techniques to predict in vivo dermal absorption (Dow Corning Corporation 1996a). Skin was excised from the dorsal surface of male and female rats, dermatomed to a split thickness of 381–629 µm, and 20 µl radiolabeled D5 applied for 24 hr. This duration was selected to replicate human usage of a product that would be washed off within 24 hr after application. Results showed that the majority of the radioactivity was volatilized and detected in the charcoal basket placed next to the skin sample (61.03%) while only 0.32% was recovered from the surface of the skin (Table 15). The rest of the skin contained 1% of applied dose and 0.38% of applied dose was found in the receptor fluid. Overall, the total mean absorbed dose of D5 based upon total radioactivity present in the skin and receptor fluid was 1.54 ± 0.43% in females and 1.08 ± 0.16% in males. Comparing results of both the in vitro and in vivo study, the total absorption of D5 was similar in both studies (Table 16) (Dow Corning Corporation 1996b; 1996b 1996a). These investigators concluded that under the conditions of both studies, approximately 1% of applied radiolabeled D5 was percutaneously absorbed, supporting the conclusion this in vitro method correlates reliably with in vivo study results.

Table 15. In vitro disposition of radioactivity following application of ¹⁴C-D5 to female and male rat skin (% applied dose ± SD).

Sex	Charcoal Basket	Skin Surface^1	Skin	24-hr Receptor Fluid	Total Absorbed^2
Female	56.19 ± 11.82	0.32 ± 0.33	1.19 ± 0.44	0.35 ± 0.18	1.54 ± 0.43
Male	69.51 ± 10.76	0.32 ± 0.14	0.67 ± 0.08	0.42 ± 0.10	1.08 ± 0.16
Average (n = 11)	61.03 ± 12.80	0.32 ± 0.27	1.00 ± 0.43	0.38 ± 0.16	1.38 ± 0.42

¹Based on the total radioactivity found in the filter, soap gauze, ethanol gauze, swab, and tape samples.

²Based on the total radioactivity found associated with the skin and in the receptor fluid.

Table 16. Comparison of in vitro¹/in vivo² absorption of ¹⁴C-D5 following dermal exposure (%applied dose).

Study/Gender	Skin (Dose Site)	Biological Samples	Total Absorption
In vitro Female	1.19 ± 0.44	0.35 ± 0.18^3	1.54 ± 0.43
In vitro Male	0.67 ± 0.88	0.42 ± 0.10^3	1.08 ± 0.16
In vivo Female	0.03 ± 0.01	0.77 ± 0.30^4	0.81 ± 0.31
In vivo Male	0.59 ± 0.81	0.28 ± 0.05^4	0.87 ± 0.83

¹(Dow Corning Corporation 1996a).

²(Dow Corning Corporation 1996b).

³Receptor Fluid.

⁴The values reported here are all model simulated.Urine, CO2, Feces, and Carcass.

In comparison, Jovanovic et al. (2008) reported lower absorption of D5 in human abdominal skin samples exposed to radiolabeled D5 either as an antiperspirant formulation or neat. Dermatomed skin samples were exposed to D5 formulations for 24 hr at an average dose of neat D5 of 6.2 mg/cm² and formulated D5 of 7.7 mg/cm². Following the 24-hr our exposure, results indicated the majority of the test article volatilized and was collected into charcoal baskets, 91 and 97% for the neat and antiperspirant formulation, respectively (Table 17). The total amount of applied D5 absorbed was 0.4% for neat D5 and 0.02% for the antiperspirant formulation. The neat solution resulted in 0.36% D5 that remained on the skin and 1.64% of D5 in the antiperspirant formulation remained. The steady state flux for neat D5 was 0.004 µg/cm²/h, while the steady state flux for formulated D5 was 0.009 µg/cm²/hr (Jovanovic et al. 2008). Permeability constants (K_p), calculated by dividing the steady state flux by the concentration of D5 in the dosing solution, were 4.2 × 10⁻⁹ cm/hr for neat D5 and 1.6 × 10⁻⁸ cm/hr for formulated D5.

Table 17. In vitro disposition of radioactivity 24 hours after application of¹⁴C-D5 to human skin¹.

Formulation	N^2	Volatilized	Skin Surface^3	Skin	Receptor Fluid	Total Absorbed^4	Recovery	Steady State Flux (µg/cm^2/hr)	Kp (cm/hr)
Neat	5	91.1 ± 1.6	0.36 ± 0.01	0.04 ± 0.01	0.002 ± 0.00	0.04 ± 0.01	91.4 ± 1.6	0.004^5	4.2×10^−9
Antiperspirant	4	97.2 ± 1.3	1.64 ± 1.11	0.02 ± 0.01	0.002 ± 0.00	0.02 ± 0.01	98.8 ± 1.2	0.009^6	1.6×10^−8

¹%applied dose ± Standard error.

²Number of individual donor skin specimens that showed acceptable skin integrity. Skin specimen from each donor was tested in duplicates.

³Total radioactivity found in extracts of swabs, tape strips, and diffusion cells.

⁴Total radioactivity associated with the skin and receptor fluid.

⁵Between 6–21 hr.

⁶Between 6–24 hr.

Dudzina et al. (2015) conducted an in vitro study to evaluate the evaporation rates of D5 from the surface of the skin following use of leave-on dermal cosmetic and PCP as reported following in vivo and in vitro application of D5, the majority of the D5 applied to the skin volatilizes prior to absorption. Therefore, Dudzina et al. (2015) investigated the evaporation rate of neat D5. A gravimetric experiment was conducted where 80 and 350 µl of D5 were applied to filter paper in a petri dish (2 replications) and placed on a scale. The sample weight was recorded every 5 min for 30 min. The gravimetric evaporation rate was determined to be approximately 0.01 mg/cm²/min. In addition, neat D5 was applied to the ear skin samples from 3 pigs at a dose of 9.5 mg. Face cream and deodorant formulations were also utilized for this study at doses of 15.4 mg/cm² and 8.6 mg/cm², respectively. The amount of D5 on or in the skin surface, in the receptor fluid, retained in the chamber and captured in the vapor trap were measured after 15, 30, 45, 60, and/or 75 min treatment. The results from the porcine skin experiments demonstrated similar evaporation rates between neat D5 and the face cream and deodorant formulations containing D5. Evaporation rates ranged from 0.056 to 0.058 mg/cm²/min. Further, for all formulations, the concentration that reached the receptor fluid was found to be negligible. Over 50% of neat D5 was observed to evaporate after 1 hr while 40% remained on the surface of the skin. A similar observation was made for the face cream. The total recovery of applied dose from the skin ranged from 77% to 100% of applied dose with a mean of 91%.

Krenczkowska et al. 2019; 2020) conducted in vitro studies with D5 focused on the ability of D5 to overcome the skin barrier and permeate into the epidermis, dermis, and receptor fluid. Dermal absorption was determined using ATR-FTIR spectroscopy and GC-FID. Human abdominal cadaver skin was extracted and 24-hr D5 dosing (100 µg/ml or 10 µg/ml) was facilitated using a static Franz type diffusion cell system, with the donor compartment of the chamber remaining closed to prevent evaporation from application site as noted by these investigators. Based upon the intensity of the specific bands observed, D5 showed the ability to penetrate the stratum corneum and permeate the epidermis and dermis. Skin protein interactions between D5 and the stratum corneum were noted producing changes in stratum corneum structure and reduction in the skin barrier (Krenczkowska et al. 2019). D5 accumulated primarily in the stratum corneum based upon the GC-FID results (Krenczkowska et al. 2019, 2020). The cumulative doses found in each layer are reported in Table 18. Permeation into the receptor fluid was also observed and indicated potential for systemic absorption.

Table 18. Cumulative dose of D5 in particular layers and fluid acceptor.

Sample^1	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2019)	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2020)
Stratum corneum	47.26	63.9
Epidermis	29.23	29.9
Dermis	11.56	13.9
Fluid Acceptor	7.32	7.4
Total Cumulative Dose	95.37	114.1

Mojsiewicz-Pieńkowska et al. (2020) used DHM to further investigate the effects of D5 on the skin barrier using human cadaver skin prepared using the same methods as employed by Krenczkowska et al. 2019; 2020). Results of DHM indicated irreversible damage of the stratum corneum which consisted of (1) destabilization of intercellular lipid lamellae and corneocyte structure, (2) loss of stratum corneal integrity and format of lacunae, and (3) alterations in the surface geometrical topography of the stratum corneum, similar to other lipophilic solvents (Barba et al. 2016; Goffin, Letawe, and Piérard 1997).

Overall, in vivo studies conducted with D5 indicated dermal absorption of D5 through the skin ranged from 0.2 to 0.8% of applied dose (Dow Corning Corporation 1996b; Jovanovic et al. 2008). Jovanovic et al. (2008) reported that ≤ 0.5% applied dose of D5 reached systemic circulation. The majority of D5 applied to the skin was found to volatilize from the skin surface before dermal absorption (Dow Corning Corporation 1996b; Jovanovic et al. 2008). Studies performed in vitro support these findings with only minimal absorption of D5 occurring in human (0%) and rat (1.38%) skin (Dow Corning Corporation 1996a; Jovanovic et al. 2008). The PBPK model from Reddy et al. (2007) simulated the available plasma and exhaled air concentrations following dermal application of D5 in the in vivo studies, with 0.05% and 0.05% of D5 predicted to reach systemic absorption in women and men, respectively. Studies using ATR-FTIR spectroscopy and GC-FID to determine the distribution of D5 at each skin layer for continuous exposures under conditions where evaporation was prevented reported > 75% of the applied D5 remained on the skin surface, which is slightly higher than % D4 on the skin surface reported by Krenczkowska et al. 2019; 2020), with the highest amount of dose absorbed through the skin noted in the epidermis (Krenczkowska et al. 2019, 2020). As with D4, irreversible damage of the stratum corneum was found following in vitro exposure to D5 (Mojsiewicz-Pieńkowska et al. 2020).

Dodecamethylcyclohexasiloxane (D6)

Limited data was identified for dodecamethylcyclohexasiloxane or D6 (CAS Registry Number 540-97-6). Four in vitro studies (Dow Corning Corporation 2003a; Krenczkowska et al. 2019, 2020; Mojsiewicz-Pieńkowska et al. 2020) were identified. Using human abdominal skin samples from 6 cadavers (5 males, one female), Dow Corning Corporation (2003a) conducted an in vitro study to determine the percutaneous absorption of neat D6 under semi-occluded conditions. A dose of 6 mg/cm² radiolabeled D6 was applied to the skin samples for a 24-hr period. Data demonstrated that the mean cumulative penetration of D6 into the receptor fluid was 0.16 ± 0.014 µg/cm² over 24 hr, and the permeability coefficient (K_p = 7.24×10⁻⁹ cm/hr) was determined based upon penetration curve determined throughout the dosage period. Steady state was achieved from 12 to 24 hr post exposure and a significant amount of the test article remained on the skin surface throughout the testing period. At the end of exposure, the majority of applied dose, approximately 46%, was found on the skin surface, while approximately 40% was volatilized from the skin surface. Only approximately 0.003% penetrated the skin into the receptor fluid with 3% detected on the skin surface. The % D6 recovered from all samples was 89%.

Krenczkowska et al. 2019; 2020) conducted in vitro studies with D6 using ATR-FTIR spectroscopy and GC-FID, to evaluate the ability of D6 to overcome the skin barrier and permeate into the epidermis, dermis, and receptor fluid. Human abdominal cadaver skin was extracted and 24-hr D6 dosing (100 µg/ml or 10 µg/ml) was facilitated using a static Franz type diffusion cell system such that evaporation of D4 might be prevented. Based upon the intensity of the specific bands observed, D6 penetrated the stratum corneum and permeated the epidermis and dermis. Skin protein interactions were found to be altered from exposure to D6 (Krenczkowska et al. 2019). The majority of D6 accumulated in the stratum corneum based on the GC-FID results (Krenczkowska et al. 2019, 2020). The cumulative doses detected in each layer are reported in Table 19. Permeation into the receptor fluid was observed.

Table 19. Cumulative dose of D6 in particular layers and fluid acceptor.

Sample (n = 5)	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2019)	Cumulative Dose (µg/cm^2/24 hr) Krenczkowska et al. (2020)
Stratum corneum	59.53	67.2
Epidermis	10.75	10.7
Dermis	5.97	5.3
Fluid Acceptor	0.94	0.90
Total Cumulative Dose	77.19	84.1

Mojsiewicz-Pieńkowska et al. (2020) used DHM, to further investigate the effects of D6 on the skin barrier using the same methods as Krenczkowska et al. 2019; 2020). Irreversible damage of the stratum corneum consisting of (1) destabilization of intercellular lipid lamellae and corneocyte structure, (2) loss of stratum cornea integrity and format of lacunae, and (3) changes in the surface geometrical topography of the stratum corneum was reported, similar to other lipophilic solvents.

While there were no in vivo studies for D6, in vitro findings indicate 0.003% D6 absorbed through the skin, 46% applied dose reported to remain on the skin, and 40% volatilized off the skin following application (Dow Corning Corporation 2003a). In addition, (studies using ATR-FTIR spectroscopy and GC-FID to determine the distribution of D6 at each skin layer found almost 80% of the dermally applied D6 dose remained on the skin surface, compared to approximately 60% for D4 and 75% for D5 (Krenczkowska et al. 2019, 2020). Krenczkowska et al. 2019; 2020) noted that, similar to D4 and D5, D6 penetrates the stratum corneum and permeates to the epidermis and dermis; however, the higher lipophilicity of D6 reduced its permeation into the deeper layers of the skin.

Linear siloxanes

The linear siloxanes considered in this review consist of the low molecular weight volatile siloxanes, hexamethyldisiloxane or L2 (CAS Registry Number 107-46-0), octamethyltrisiloxane or L3 (CAS Registry Number 107-51-7), decamethyltetrasiloxane or L4; (CAS Registry Number 141-62-8), dodecamethylpentasiloxane or L5 (CAS Registry Number 141-63-9), and the high molecular weight polydimethylsiloxane or PDMS (CAS Registry Number 9016-00-6, 63148-62-9, and 63,394-02-5). Linear siloxanes have the chemical formula [R₃SiO]_n and are named according to the number of repeated units. A limited number of studies, summarized below, were performed to investigate dermal absorption of linear siloxanes and no apparent data were identified specifically for L3 or L5. Linear siloxanes, either in the pure form or as emulsions, have various uses in a wide variety of industrial and consumer products including PCP and medicinal products where dermal exposure may be of interest (ECETOC 2011).

Polydimethylsiloxane (CAS RN 9016-00-6, 63148-62-9, and 63,394-02-5)

While there are a large number of linear siloxane compounds available, the most common is PDMS, a highly flexible silicone with a low boiling point and low viscosity (EFSA 2015). PDMS may be a viscous liquid or an elastic solid. The chemical structure of PDMS is composed of repeating siloxy units [(CH₃)₂SiO], best represented by the main CAS Registry Number [63148-62-9], although other CAS RNs ([9016-00-6] and [63394-02-5]) may be considered representative of this class of compounds (ECETOC 2011). PDMS compounds are clear, colorless, odorless, viscous fluid at normal ambient temperature. Viscosity depends on and increases with the number of repeating siloxy units and may range from 10 to > 100,000 centistokes (cSt).

In the Cosmetics Ingredient Review’s final report on the safety assessment of dimethicone (Nair and Cosmetic Ingredients Review Expert Panel 2003), one study was included that focused on the dermal absorption of PDMS (Hobbs, Fancher, and Calandra 1972). In this study, dimethicone (100 cSt) was applied to the backs of 5 male subjects once daily for 10 days at a daily dose of 50 mg/kg bw. The test material was evenly distributed over the entire back with no covering. After 20 hr exposure each day, any excess material was rinsed off. The absorption of the test substance was measured as the amount of total elemental silicon in the blood and urine on days 1, 3, 6, 8, and 10. Data demonstrated no significant changes in blood or urine silicon concentrations in any subjects compared to baseline measurement prior to dosing, indicating no marked evidence of dermal absorption of dimethicone.

ECETOC (2011) reports on two unpublished studies, one in vivo study in rats (Plotzke, McMahon, and Hubbell 1994) and an in vitro study in human abdominal skin and vaginal tissue (Plotzke, Andriot, and Naas 2000) conducted by Dow Corning Corporation. In the in vivo study conducted in rats, radiolabeled PDMS of 350 cSt viscosity, was applied under occlusive contact to the dorsal surface of male rats. The animals were maintained in metabolism cages for 24 hr after which the exposure site was washed and re-wrapped with fresh non-occlusive bandages. The rats were returned to the metabolism cages for urine, feces, and expired air collection for an additional 72 hr. Following this period, the animals were killed, and exposure site excised. Results showed that approximately 70% of the administered dose was found on the skin surface and 11.4% was in the skin at the site of application. Low levels were detected in the feces (0.01%) and the CO₂ traps (0.001%). Results indicated little dermal absorption of 350 cSt PDMS with no radioactivity detected in the blood.

In the in vitro study reported in ECETOC (2011), infinite doses of 10 and 350 cSt radiolabeled PDMS (10 mg/cm³) were applied for 96 hr to human abdominal skin and vaginal tissue mounted on Franz in vitro diffusion cells. The dermal flux rates for the 350 cSt PDMS was 0.3 and 2 ng/cm²/hr for abdominal skin and vaginal tissue, respectively. For the 10 cSt PDMS, the dermal flux rates were 0.2 and 6 ng/cm²/hr for abdominal skin and vaginal tissue, respectively. Therefore, based upon dermal flux, dermal absorption of 10 cSt and 350 cSt is approximately an order of magnitude greater in vaginal tissue compared to abdominal skin. However, ≤ 0.5% of the applied dose was bioavailable regardless of the tissue or viscosity of the PDMS. Overall, the in vivo and in vitro dermal absorption studies conducted with PDMS all indicate little to no systemic absorption of PDMS.

Low molecular weight volatile linear siloxanes

No apparent data were identified specifically for L3 or L5. The dermal penetration of the linear siloxane, hexamethyldisiloxane (HMDS or L2) was studied in vitro using human cadaver skin (Dow Corning Corporation 2000a). Human abdominal skin samples were acquired from 6 cadavers (5 males, one female) approximately 24 hr after death. Full thickness skin samples were placed in a Bronaugh Flow-Through Percutaneous apparatus and the target HMDS dose was 9 mg/cm² (actual dosages ranged from 7.0 to 10.3 mg/cm²). After 24 hr, 0.023 ± 0.011% of the applied dose of neat HMDS was absorbed, with the majority absorbed dose (0.017% ± 0.01) found in the skin. The amount of applied dose recovered was 97.58 ± 0.35% with the majority of the test article volatilized off the skin surface and recovered in charcoal traps (97.51 ± 0.357%). Following the 24-hr diffusion period, the cumulative penetration of HMDS was 0.352 ± 0.037 µg/cm². The steady state flux was 0.0356 µg HMDS/cm²/hr, and the permeability coefficient (Kp) was calculated to be 4.66 × 10⁻⁸ cm/hr. Overall, as reported for the cyclic siloxanes, the majority of HDMS volatilized from the skin surface. The small amount of HDMS absorbed was detected in the skin and did not penetrate into the receptor fluid.

An in vitro study was conducted to determine the dermal penetration of decamethyltetrasiloxane (L4) on human skin under semi-occluded conditions (Dow Corning Corporation 2006b). Using the Bronaugh flow-through diffusion cell system, human abdominal skin samples were exposed to a target application dose of 10 mg¹⁴C-L4/cm² (actual dose was 9.3 mg/cm²). Almost all of applied dose (99.9%) was volatilized from the skin surface and recovered in charcoal baskets above the exposure site. A small amount (0.06% of applied dose) was recovered on the skin surface after 24 hr treatment, 0.03% was present in skin after washing and tape stripping. 0.001% of the applied dose penetrated through the skin into the receptor fluid with cumulative penetration over 24 hr determined to be 0.07 ± 0.01 µg/cm². The permeability coefficient (Kp) was calculated to be 3.9 × 10⁻⁹ cm/hr. Results of the study indicated almost all of the recovered L4 was volatilized from the skin surface and the total % absorbed was determined to be 0.03% of the applied dose.

Conclusions and discussion

There are currently more than 150,000 practical applications of cyclic and linear siloxanes including uses in the pharmaceutical, medical, cosmetic and food production industries (Mojsiewicz-Pieńkowska et al. 2016). It has been reported that over half of the skin PCP marketed to consumers contain at least one type of siloxane (Mojsiewicz-Pieńkowska et al. 2016).; Due to the use of both cyclic and linear siloxanes in consumer products with the potential for human dermal exposure this review focused specifically on the identification and review of the available data associated with the dermal absorption/penetration and fate of some of the most commonly used cyclic siloxanes, D4, D5, and D6; and linear siloxanes, L2, L3, L4, L5, and PDMS.

While the permeability of human skin differs from animal skin (Zareba et al. 2002), the results of studies conducted in humans, rats and ex vivo human skin demonstrate very low absorption of cyclic siloxanes, D4, D5, and D6; and formulations containing D4 and/or D5 following in vitro and in vivo dermal application. For all of the cyclic siloxanes, the majority of substance applied to the skin was reported to volatilize from the skin surface. In humans, Biesterbos et al. (2015) concluded that dermal exposure to neat D4 or D5; or deodorant and night cream formulations containing both compounds contributed minimally to dermal uptake, and due to volatilization from the skin surface, inhalation is likely the primary route of exposure.

Overall, the results of the in vivo studies conducted in humans and rats dermally exposed to D4 or D5 indicate that ≤ 1% of the dermally applied dose of D4 or D5 is absorbed through the skin, with ≤ 0.5% of applied dose passing through the skin to systemic compartments (Dow Corning Corporation 1996b; Jovanovic et al. 2008; Zareba et al. 2002). The majority of D4 applied to the skin volatilizes from the skin surface before dermal absorption (Biesterbos et al. 2015; Jovanovic et al. 2008; Zareba et al. 2002). Studies performed in vitro support these findings with minimal absorption of D4 occurring (0.5% to 1%) in human skin and < 0.05% in swine skin (Jovanovic et al. 2008; Zareba et al. 2002). Similarly, results of in vitro studies with D5 report 0.4% absorption in human skin (Jovanovic et al. 2008) with slightly higher absorption noted in rat skin (1.38%) (Dow Corning Corporation 1996a). While no in vivo studies for D6 were identified, in vitro results indicate D6 dermal absorption is less than D4 or D5, with 0.003% reported to be absorbed through the skin (Dow Corning Corporation 2003a). Results of all in vivo studies conducted in both humans and animals concluded the majority of D4 or D5 applied to the skin surface volatilized prior to dermal absorption.

In vitro skin distribution study results for D4, D5 and D6 using ATR-FTIR spectroscopy and GC-FID indicated approximately 60 to 80% of the dermally applied dose of D4, D5, and D6 remains on the skin surface (Krenczkowska et al. 2019, 2020), while most of the dose absorbed through the skin surface was found in the epidermis. While D4 and D5 were also identified in the deeper layers of the skin and adipose tissue; the lipophilic properties of D6 prevented the substance from permeating in large amounts through the deeper layers of skin. As noted following in vivo dermal exposure in humans and animals, the majority of D4, D5, or D6 applied to the skin surface volatilized prior to dermal absorption.

Reddy et al. (2007) modified the D4 and D5 inhalation models to incorporate the dermal route of exposure. These models were then applied to simulate the results of in vivo studies with human volunteers. This analysis determined that 0.3 and 0.12% of D4 reached systemic circulation in women and men, respectively. Similar results were estimated for D5, with 0.05 and 0.05% of D5 predicted to reach systemic circulation in women and men, respectively. Model simulations also estimated that of the D4 absorbed into the skin compartment, almost all (>99%) moved back to the skin surface and evaporated following the exposure. Jovanovic et al. (2008) also noted that the largest amount of the absorbed dose (>90%) was present in the skin and had not permeated through the skin layers and into the receptor fluid.

Few studies investigating the dermal absorption of linear siloxanes were located. For PDMS, one in vitro study and two in vivo studies, one in humans and one in rats were identified. No data were identified specifically for L3 and L5, and one in vitro study each have been conducted for L2 and L4. Overall, the in vivo and in vitro dermal absorption studies conducted with PDMS, L2 and L4 all indicate little to no systemic absorption of PDMS with the majority of the applied dose volatilizing from the skin surface. For both the cyclic and linear siloxanes that are the focus of this review, experimental data all support a low level of dermal absorption. Mojsiewicz-Pieńkowska et al. (2020) found the potential for damage of the stratum corneum after prolonged skin contact with liquid D4, D5, or D6 under experimental conditions where evaporation was prevented. Prolonged exposure of the skin to liquid solvents is frequently associated with damage to the stratum corneum (Barba et al. 2016; Goffin, Letawe, and Piérard 1997), but these effects would not be expected for the siloxanes under normal exposure conditions where evaporation from the skin is not prevented (e.g., body lotions, foundations), due to the volatility of both the cyclic and linear siloxanes. Various investigators demonstrated that siloxanes applied to the skin rapidly volatilize, preventing prolonged contact (Dow Corning Corporation 2000a, 2006b, 2003b, 2003a, Jovanovic et al. 2008). In adults and children, the use of PCP applied to the skin such as body lotions or diaper creams provide the highest contribution of D4 and D5 dermal exposure, all of which are applied by nonoccluded manner enabling volatilization and evaporation from the skin surface (Franzen et al. 2017; Gentry et al. 2017). The application rate of consumer products containing D4 and D5 ranged from 0.06 g/day for cosmetic night cream to approximately 14 g/day in hair conditioner (Gentry et al. 2017).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, RG, upon reasonable request.

Additional information

Funding

This work was supported by the Silicones Environmental, Health and Safety Center (SEHSC).